Overview

• Data Scavenger Hunts
• Example: NHANES
• The burro R package: make a dataset explorable

Motivation

• Students really don’t look at data
• Don’t see Exploratory Data Analysis as valuable

What are Data Scavenger Hunts?

• Centered around a dataset
• Group/Collaborative Activity
• Web app for exploring data
• Questions to explore and report back

Example: NHANES

• National Health and Nutrition Examination Survey
• Lots of outcomes, and is of broad interest
• Looked at Depression, Type 2 Diabetes, Physical Activity as outcomes

Example Questions

• Is there an association of LittleInterest with Depressed?
• Is marijuana use associated with depression?
• What is the oldest person in the dataset? Any idea why?

Burro(w) into the data

Is LittleInterest associated with Depressed?

categorical exploration

Burro: Boxplots

Are Sleep Hours associated with Depression?

## Burro package

• R package that builds shareable Shiny apps for data exploration
• Authors: Jessica Minnier and Ted Laderas
• http://laderast.github.io/burro/

burro

Outcomes

• “The scavenger hunt. It is great place to apply this knowledge.”
• “I enjoyed the collaborative and self-paced approach.”
• “I liked the dataset and the open curiosity element of the workshop.”

Noticable improvement in confidence to do Data Exploration!

For more info

• Slides are here: http://bit.ly/lst_talk
• laderast@ohsu.edu
• Try it out! http://tladeras.shinyapps.io/nhanes_explore/